Musculoskeletal Imaging Original Research

Transcription

1 Musculoskeletal Imaging Original Research Lee et al. MRI of Anterior Talofibular Ligament Injury Musculoskeletal Imaging Original Research Min Hee Lee 1 Jang Gyu Cha 1 Young Koo Lee 2 Gyo Chang Choi 3 Sang Hyun Paik 1 Hae Kyung Lee 1 Seong Jin Park 4 Hyun Joo Kim 5 Lee MH, Cha JG, Lee YK, et al. Keywords: arthroscopy, ligament injury, MRI DOI: /AJR Received March 16, 2011; accepted after revision September 5, Department of Radiology, Soonchunhyang University Bucheon Hospital, 1174 Jungdong, Wonmi-Gu, Bucheon-Si, Gyunggi-Do , Korea. Address correspondence to J. G. Cha (mj4907@schmc.ac.kr). 2 Department of Orthopedics, Soonchunhyang University Bucheon Hospital, Gyunggi-Do, Korea. 3 Department of Radiology, Soonchunhyang University Gumi Hospital, Gumi-si, Gyeongbuk, Korea. 4 Department of Radiology, Kyung-hee University Medical Center, Hoegi-dong, Dongdaemun-Gu, Seoul, Korea. 5 Department of Radiology, Soonchunhyang University Seoul Hospital, Seoul, Korea. AJR 2012; 198: X/12/ American Roentgen Ray Society The Bright Rim Sign on MRI for Anterior Talofibular Ligament Injury With Arthroscopic Correlation OBJECTIVE. The purpose of this article is to determine whether bright rim lesions on MRI are a marker for anterior talofibular ligament injury. MATERIALS AND METHODS. The study included 34 patients who had an ankle injury and underwent arthroscopic surgery. All patients underwent 3-T MRI for the diagnosis of anterior talofibular ligament injury. If MRI revealed nonvisualization of the ligament, ligament discontinuity, and unusual ligament thickening (criterion 1) or the bright rim sign (criterion 2), the injury was considered to be a ligament disruption. After MRI, ankle arthroscopy was performed in all patients for a definitive diagnosis. RESULTS. Arthroscopy showed anterior talofibular ligament disruption in 33 patients. When the MRI diagnosis was based on criterion 1, anterior talofibular ligament disruption was diagnosed with a sensitivity of % and an accuracy of %. When the MRI diagnosis was based on both criteria 1 and 2, anterior talofibular ligament disruption was diagnosed with a sensitivity of % and an accuracy of %. By adding criterion 2 to the diagnosis, the sensitivity for anterior talofibular ligament injury was increased significantly (p < 0.01), and 8 12 additional patients with anterior talofibular ligament injury were diagnosed, most of whom exhibited a partial tear of the anterior talofibular ligament on arthroscopy. The interobserver agreement rate for the presence of anterior talofibular ligament disruption using criterion 1, both criteria, and the bright rim sign was fair to excellent. CONCLUSION. A cortical defect with bright dotlike or curvilinear high-signal-intensity lesions on T2-weighted MRI may be an additional morphologic feature to increase the diagnostic performance of detecting anterior talofibular ligament injuries, including those with partial tears. T he anterior talofibular ligament is the most commonly injured ligament in ankle inversion injuries [1, 2]. According to biomechanical studies, the anterior talofibular ligament is the weakest lateral ankle ligament, followed by the calcaneofibular ligament [3]. Although most ankle sprains are treated conservatively with functional rehabilitation or an Aircast or Air-Stirrup ankle brace (DJO) [4] for persistent or residual ankle disability, surgical approaches, such as reapproximation of the torn ligament or reconstruction of the ligament repair may be considered as treatment options [5, 6]. In these situations, because of the advantage of noninvasiveness and superior soft-tissue imaging capabilities, MRI may be a useful diagnostic modality to delineate intraarticular or extraarticular sources of ankle pain after a sprain that might not have been apparent clinically [7 12]. The anterior talofibular ligament courses obliquely, posterolaterally to anteromedially with respect to the tibiotalar joint, which allows resistance of internal rotation of the talus and not just anterior translation [13]. On MRI, the anterior talofibular ligament is normally observed as a linear structure with low signal intensity on T1- and T2-weighted images. Several investigators have previously defined the characteristic MRI appearance of anterior talofibular ligament injury as ligament discontinuity, wavy or curved contours, elongations, contour irregularities, and increased signal intensity within the ligament on protonor T2-weighted images [7, 10, 14]. Although such findings have been thought to be sufficiently distinctive, obtaining reliable reproducible MRI examinations of ankle ligament abnormalities remains challenging [13]. We have recently noticed a characteristic feature of anterior talofibular ligament injury AJR:198, April

2 Lee et al. as depicted on MRI, a feature that we term the bright rim sign. To our knowledge, however, no previous report has described a bright signal intensity focus at the talar or fibular attachment site of the anterior talofibular ligament. The aim of this study was to determine whether the bright rim sign on MRI is a reliable marker of anterior talofibular ligament injury in patients with lateral ankle disability. Materials and Methods Patients We evaluated all patients who underwent ankle MRI because of ankle sprains or ankle instability between March 2009 and February Among 74 patients who had undergone ankle MRI, 48 who underwent arthroscopy were included in this study. Among the latter, 14 were excluded because of inadequate medical records (n = 12), arthroscopy was performed before the MRI (n = 1), or the interval was too long (> 3 months) between the MRI and arthroscopic surgery (n = 1). Thus, 34 patients were enrolled (22 male and 12 female patients; age range, years; mean age, 29 years) (Fig. 1). Ethics review board approval and informed consent were not required for this study because it involved only retrospective image or case record review. MRI MRI examinations were performed using a 3-T MRI system (Signa HDx, GE Healthcare) and a 20-cm extremity coil. The feet of the subjects were placed in the neutral position. Axial, sagittal, and coronal images were obtained to detect disorders of the ankle joint. To evaluate the anterior talofibular ligament, axial planes were used primarily. The plane examination protocol consisted of transverse T2-weighted fast spin-echo sequences with the following parameters: TR/TE, 4000/55; matrix, ; echo-train length (ETL), 16; number of signals acquired, 2; and FOV, 140 mm. For coronal fat-saturated T2-weighted images, the parameters were as follows: TR/TE, 4000/70; ETL, 12; FOV, 140 mm; image matrix, ; slice thickness, 3 mm; and interslice gap, 0.3 mm. For sagittal fat-saturated T2-weighted images, parameters were as follows: TR/TE, 4000/55; ETL, 18; FOV, 140 mm; image matrix, ; slice thickness, 3 mm; and interslice gap, 0.3 mm. Image Interpretation After data collection for all 34 patients, the MRI scans were independently reviewed by three experienced musculoskeletal radiologists (with 10, 5, and 1 years of experience, respectively), who were blinded to the patients histories, outcomes, and operative findings. Ankle MRI because of ankle injury (n = 74) Underwent arthroscopy (n = 48) Enrolled patient (n = 34) (age range, years; mean, 29 years) When MRI revealed any feature of diagnostic criterion 1 (i.e., nonvisualization of the ligament, discontinuity of the ligament, wavy or curved contour, or increased signal intensity within the ligament), the injury was considered to be a ligament disruption. A dotlike or curvilinear high signal intensity with cortical disruption appearing on the fibular or talar attachment site of the anterior talofibular ligament on axial T2-weighted images was referred to as the bright rim sign. The bright rim sign was used as diagnostic criterion 2. In the first review session, each reviewer independently examined the MRI examinations and noted whether an anterior talofibular ligament injury existed according to criterion 1 alone. In the second review session, each reviewer noted whether a bright rim sign was identified and its location. The diagnosis of ligament injury was made after fulfillment of either criterion 1 or 2. There was a 2-week interval between each image review session to eliminate recall bias. Ankle Arthroscopy Ankle arthroscopy was used as the standard of reference in this study. Before ankle arthroscopy, we provided the arthroscopist with the MRI reports, which did not include information regarding the bright rim sign. Patients were examined by ankle arthroscopy before the operation. Arthroscopic studies were considered to be positive for the presence of at least one of the following diagnostic criteria for determining an anterior talofibular ligament injury: an abnormal course of the ligament, a decrease in the tautness of the ligament, discontinuity of the ligament with or without the defect being filled by fibrous tissue, and an avulsion at the attachment of the fibula or talus [7, 8, 15]. In addition, the surgeon noted whether the ligament No arthroscopy (n = 26) Inadequate record (n = 12) Arthroscopy before MRI (n = 1) Too long interval between MRI and arthroscopy (n = 1) Fig. 1 Flow diagram of patient exclusions leading to selection of final study group. was disrupted completely or partially. Ligament lesions were graded as a partial tear if the ligament was thinned or elongated or both and as a complete tear if the continuity was lost, with or without the defect being filled by fibrous tissue, or if there was avulsion at the attachment of the fibula or talus [16]. Statistical Evaluation The arthroscopic result was considered to be the standard of reference. MRI results were compared with arthroscopic results. Descriptive statistics were used to compare the results between using MRI criterion 1 alone and both MRI criteria. Statistically significant differences in sensitivity between using diagnostic criterion 1 alone and using both criteria 1 and 2 were calculated using the McNemar test. Statistical significance was set at p less than A kappa statistic was recorded for interobserver agreement regarding the presence and location of the bright rim sign on MRI. Interobserver agreement rates for the presence of anterior talofibular ligament disruption in each review session were also estimated. The results of the kappa test were interpreted as follows: kappa values of indicated poor correspondence, indicated fair correspondence, indicated good correspondence, and indicated excellent correspondence [17]. All statistical operations were processed using the Stata statistical software (version 11.0, StataCorp). Results The interval between the MRI and arthroscopic surgery ranged from 1 to 84 days (median, 9 days; mean 14 days). According to arthroscopic findings, 33 of 34 patients had an- 886 AJR:198, April 2012

3 MRI of Anterior Talofibular Ligament Injury A Fig year-old man with complete tear of anterior talofibular ligament of left ankle. A, Axial T2-weighted MRI scan reveals tiny dotlike (arrow) and curvilinear (arrowhead) high-signal-intensity lesions at anterior aspect of distal fibula and talus. This constitutes bright rim sign. B, Axial T2-weighted MRI scan at slightly lower level than in A shows complete detachment (arrow) of anterior talofibular ligament from distal fibula. C, Arthroscopic photograph of anterior talofibular ligament (ATFL) shows complete disruption of ligament (arrow) at fibular attachment site. F = fibula, T = talus. A Fig year-old man who complained of right lateral ankle instability. Diagnosis was partial tear of anterior talofibular ligament. A and B, Sequential axial T2-weighted MRI scans show diffuse thickening of anterior talofibular ligament, with increased signal intensity (arrowheads, A) and dotlike high-signal-intensity lesion (arrow, A and B) in subcortical bone marrow and cortical disruption. C, Arthroscopic findings show focal hypertrophy (arrow) of anterior talofibular ligament (ATFL) at fibular attachment. AITFL = anterior inferior tibiofibular ligament, F = fibula, T = talus. terior talofibular ligament disruption. They included 11 patients with a complete disruption and 22 with a partial disruption. One patient had an intact anterior talofibular ligament. Using arthroscopic findings as the reference standard, criterion 1 showed 22 (66.7%; readers 1 and 2) and 20 (60.6%; reader 3) of the 33 anterior talofibular ligament disruptions. The false-negative cases determined by reader 1 were confirmed as 10 partial tears and one complete tear; those determined by reader 2 were confirmed as 10 partial tears and one complete tear; and those determined by reader 3 were confirmed as 12 partial tears and one complete tear. The truepositive cases for readers 1 and 2 consisted of 10 cases with complete disruption (Fig. 2) and 12 cases with partial disruption (Fig. 3); those for reader 3 included 10 complete and 10 partial tears. For readers 1, 2, and 3, the respective sensitivities were 66.7%, 66.7%, and 60.6%; the respective specificities were 0%, 100%, and 0%; and the respective accuracies were 61.8%, 67.6%, and 58.8%. B B On axial T2-weighted MRI, the bright rim sign was observed in 20 of 34 patients for readers 1 and 2 and in 25 of 34 patients for reader 3. Bright rim lesions were observed at fibular attachment sites in four cases for reader 1, seven cases for reader 2, and five cases for reader 3; at talar attachment sites in two cases for reader 1, four cases for reader 2, and six cases for reader 3; and at both anterior talofibular ligament attachment sites in 14 cases for readers 1 and 3 and in nine cases for reader 2. The bright rim lesion was not C C AJR:198, April

4 Lee et al. A Fig year-old woman with left lateral ankle instability. A, Axial T2-weighted MRI scan shows well-visualized anterior talofibular ligament without signal change or thickening. B, Axial T2-weighted MRI scan obtained in more cranial section than in A shows tiny dotlike bright signal intensity (arrow) at fibular attachment site of anterior talofibular ligament. C, Arthroscopic findings reveal fraying (arrows) of anterior talofibular ligament (ATFL) without discontinuity. F = fibula, T = talus. TABLE 1: Statistical Analysis of Results of Three Readers observed by any reader on coronal or sagittal fat-saturated T2-weighted images. By adding criterion 2 to the diagnosis, reader 1 diagnosed eight additional cases with anterior talofibular ligament disruption, reader 2 diagnosed nine additional cases, and reader 3 diagnosed 12 additional cases. Most of the additional cases were confirmed as partial anterior talofibular ligament disruptions at arthroscopy, except for one complete tear case (Fig. 4). Using both criteria 1 and 2, the respective statistical values for readers 1, 2, and 3 were sensitivities of 90.9%, 93.9%, and 96.9%; specificities of 0%, 100%, and 0%; and accuracies of 88.2%, 94.1%, and 94.1%. For all readers, the sensitivity for diagnosing an anterior talofibular ligament tear was significantly improved (p < 0.05) by using both criteria 1 and 2 (Table 1). The interobserver agreement regarding the presence of anterior talofibular ligament disruption using both criteria 1 and 2 was fair to good (κ = ) (Table 2). The interobserver agreement rate for the presence of anterior talofibular ligament disruption using criterion 1 was excellent (κ = ) among the three reviewers. The interobserver agreement rate regarding the presence and location of the bright rim sign was fair to good (κ = ). Discussion In most cases, the diagnosis of an anterior talofibular ligament injury is heavily dependent on physical examination, radiographic stress tests, and arthrography [18]. The clinical stress test is a useful method to assess joint laxity caused by ligament injury [19]. However, the use of clinical stress tests is restricted in patients with an acute ankle sprain. Like clinical stress tests, the efficacy of stress radiographs in acute ankle sprains is limited because they are influenced considerably by other factors, such as patient pain, edema, muscle spasms, radiographic technique, and differences in the amount of force applied to the joint [7, 20]. Ankle arthrography is an alternative diagnostic method for assessing ligament injury, but it is also an invasive examination. Additionally, it provides its best B result only within 48 hours after the trauma. Furthermore, these examinations can only indirectly depict injured ligaments [7]. Several studies have documented that MRI is well suited to reveal the anatomy as well as the extent of injuries of lateral collateral ligaments [7, 8, 10, 14, 21]. In patients with an ankle injury, we have recognized the common appearance on the talus and fibula located adjacent to the expected attachment site of the anterior talofibular ligament on axial T2-weighted MRI. We have called this the bright rim sign. By using three reviewers in our study, we could address interobserver agreement in the recognition of bright dotlike or curvilinear lesions on the bony cortex and the application of diagnostic criteria described in the literature with and without additional criteria. In our study, the bright rim sign, when used with well-known diagnostic criteria for anterior talofibular ligament injury, is a reliable indicator that anterior talofibular ligament injury has occurred. In addition, interobserver agreement in recognizing this Reader 1 Reader 2 Reader 3 Parameter Criterion 1 Criteria 1 and 2 Criterion 1 Criteria 1 and 2 Criterion 1 Criteria 1 and 2 Accuracy 61.8 (21/34) 88.2 (30/34) 67.6 (23/34) 94.1 (32/34) 58.8 (20/34) 94.1 (32/34) Sensitivity 66.7 (22/33) 90.9 (30/33) a 66.7 (22/33) 93.9 (31/33) a 60.6 (20/33) 96.9 (32/33) a Specificity 0.0 (0/1) 0.0 (0/1) 100 (1/1) 100 (1/1) 0.0 (0/1) 0.0 (0/1) Note Data are % (no./total). Criterion 1 refers to nonvisualization of the ligament, discontinuity of the ligament, wavy or curved contour, and increased signal intensity within the ligament. Criterion 2 refers to dotlike high-signal-intensity lesion with no adjacent bony cortex. a Significantly different at p < 0.05 (McNemar test). C 888 AJR:198, April 2012

5 MRI of Anterior Talofibular Ligament Injury TABLE 2: Interobserver Agreement Between Readers for Image Interpretation Basis of Interpretation Reader 1 Versus 2 Reader 1 Versus 3 Reader 2 Versus 3 Presence of a bright rim Criterion Criteria 1 and Note Data are kappa values, where values of indicate poor correspondence, indicate fair correspondence, indicate good correspondence, and indicate excellent correspondence. Criterion 1 refers to nonvisualization of the ligament, discontinuity of the ligament, wavy or curved contour, and increased signal intensity within the ligament. Criterion 2 refers to dotlike high-signal-intensity lesion with absence of adjacent bony cortex. sign is high. We suggest that, because of cortical disruption induced by a small avulsion injury at the fibular or talar attachment of the anterior talofibular ligament or a kissing contusion occurring at the tibiotalar joint [22], the subcortical fat marrow was exposed to joint fluid, resulting in a chemical-shift artifact that presented as dotlike or curvilinear high-signal-intensity lesion on T2-weighted images. These lesions were not observed on fat-suppressed coronal or sagittal T2- weighted images. Bright rim lesions were observed in of 34 patients with anterior talofibular ligament injury ( %). Most of the bright rim lesions on MRI were observed at the attachment of the distal fibula and talus. We postulate that this may be attributable mainly to repeated trauma, resulting in a partial avulsion injury at both attachment sites. As noted in our results, the 8 12 ankle cases ( %) that were interpreted as normal on MRI using criterion 1 showed dotlike or curvilinear high-signal-intensity lesions, which contributed to improved diagnostic performance for detecting anterior talofibular ligaments. The reported sensitivity for diagnosing anterior talofibular ligament injury on the basis of these MRI findings ranged from 97% to 100% [7, 8, 10]. In contrast, our results using diagnostic criterion 1 showed low sensitivity ( %). This difference is not surprising because MRI has the potential to exhibit a partial volume effect caused by the use of a 3-mm section thickness, rendering subtle morphologic lesions such as narrowing or elongation of the anterior talofibular ligament difficult to detect [8]. However, we think that the use of a thinner section or a matrix with higher spatial resolution would improve the diagnostic performance of anterior talofibular ligament injury. In our study, almost all of the false-negative lesions ( %) on MRI using criterion 1 were partial tears on arthroscopy. In addition, on MRI, it is impossible to assess a decrease in the tautness of the ligament, which is an arthroscopic criterion of anterior talofibular ligament injury. When using both criteria 1 and 2, the sensitivity was %, which was comparable to that of previous qualitative studies. A limitation of this study was that subjective interpretation of cross-sectional imaging studies was required to determine the presence of a specific finding. Despite this limitation, the high degree of observer agreement for the bright rim sign suggests that this sign is a reliable finding and can be recognized with confidence. Another concern regarding the study design was that we did not include a control group without complaints and symptoms of instability; it only included patients who underwent ankle MRI and arthroscopy, which could introduce selection bias and possibly explain the very low number of falsepositive lesions in our study. However, in an attempt to solve this problem, it is both practically and ethically impossible to perform arthroscopy for healthy volunteers. Finally, when using both criteria 1 and 2, the interobserver agreement for the diagnosis of anterior talofibular ligament disruption was fair to good, which was lower than that when using criterion 1 alone. However, the actual agreement rates for criteria 1 and 2 among three readers were over 0.94, which was not described in the results. This discrepancy between high agreement rate and low kappa value for criteria 1 and 2 can be explained by the fact that, in most cases, all readers interpreted as positive for anterior talofibular ligament tear with using criteria 1 and 2. Therefore, the dominance of positive diagnosis interpreted by all readers resulted in lowering kappa value for criteria 1 and 2 [23, 24]. In conclusion, a cortical defect with a bright dotlike or curvilinear high-signal-intensity lesion on T2-weighted MRI may be regarded as an additional finding to help increase the diagnostic performance for detecting anterior talofibular ligament injuries, including partial tears. References 1. Campbell SE, Warner M. MR imaging of ankle inversion injuries. Magn Reson Imaging Clin N Am 2008; 16: Tochigi Y, Yoshinaga K, Wada Y, Moriya H. Acute inversion injury of the ankle: magnetic resonance imaging and clinical outcomes. Foot Ankle Int 1998; 19: Siegler S, Block J, Schneck CD. The mechanical characteristics of the collateral ligaments of the human ankle joint. Foot Ankle 1988; 8: Boyce SH, Quigley MA, Campbell S. Management of ankle sprains: a randomised controlled trial of the treatment of inversion injuries using an elastic support bandage or an Aircast ankle brace. Br J Sports Med 2005; 39: Javors JR, Violet JT. Correction of chronic lateral ligament instability of the ankle by use of the Brostrom procedure: a report of 15 cases. Clin Orthop Relat Res 1985; 198: Baltopoulos P, Tzagarakis GP, Kaseta MA. Midterm results of a modified Evans repair for chronic lateral ankle instability. Clin Orthop Relat Res 2004; 422: Oae K, Takao M, Uchio Y, Ochi M. Evaluation of anterior talofibular ligament injury with stress radiography, ultrasonography and MR imaging. Skeletal Radiol 2010; 39: Takao M, Innami K, Matsushita T, Uchio Y, Ochi M. Arthroscopic and magnetic resonance image appearance and reconstruction of the anterior talofibular ligament in cases of apparent functional ankle instability. Am J Sports Med 2008; 36: Delfaut EM, Demondion X, Boutry N, Cotten H, Mestdagh H, Cotten A. Multi-fasciculated anterior talo-fibular ligament: reassessment of normal findings. Eur Radiol 2003; 13: Kreitner KF, Ferber A, Grebe P, Runkel M, Berger S, Thelen M. Injuries of the lateral collateral ligaments of the ankle: assessment with MR imaging. Eur Radiol 1999; 9: Cardone BW, Erickson SJ, Den Hartog BD, Carrera GF. MRI of injury to the lateral collateral ligamentous complex of the ankle. J Comput Assist Tomogr 1993; 17: Verhaven EF, Shahabpour M, Handelberg FW, Vaes PH, Opdecam PJ. The accuracy of three-dimensional magnetic resonance imaging in the diagnosis of ruptures of the lateral ligaments of the ankle. Am J Sports Med 1991; 19: Watson AD. Ankle instability and impingement. Foot Ankle Clin 2007; 12: Rosenberg ZS, Beltran J, Bencardino JT. From the AJR:198, April

6 Lee et al. RSNA Refresher Courses: Radiological Society of North America MR imaging of the ankle and foot. RadioGraphics 2000; 20(Spec No):S153 S Frey C, Bell J, Teresi L, Kerr R, Feder K. A comparison of MRI and clinical examination of acute lateral ankle sprains. Foot Ankle Int 1996; 17: Hintermann B, Boss A, Schafer D. Arthroscopic findings in patients with chronic ankle instability. Am J Sports Med 2002; 30: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: Marder RA. Current methods for the evaluation of ankle ligament injuries. Instr Course Lect 1995; 44: Boruta PM, Bishop JO, Braly WG, Tullos HS. Acute lateral ankle ligament injuries: a literature review. Foot Ankle 1990; 11: Raatikainen T, Putkonen M, Puranen J. Arthrography, clinical examination, and stress radiograph in the diagnosis of acute injury to the lateral ligaments of the ankle. Am J Sports Med 1992; 20: Perrich KD, Goodwin DW, Hecht PJ, Cheung Y. Ankle ligaments on MRI: appearance of normal and injured ligaments. AJR 2009; 193: Sijbrandij ES, van Gils AP, Louwerens JW, de Lange EE. Posttraumatic subchondral bone contusions and fractures of the talotibial joint: occurrence of kissing lesions. AJR 2000; 175: Crewson PE. Reader agreement studies. AJR 2005; 184: Feinstein AR, Cicchetti DV. High agreement but low kappa. Part I. The problems of two paradoxes. J Clin Epidemiol 1990; 43: FOR YOUR INFORMATION The AJR has made getting the articles you really want really easy with an online tool, Really Simple Syndication, available at It s simple. Click the RSS button located in the menu on the right side of the page. You ll be on your way to syndicating your AJR content in no time. 890 AJR:198, April 2012